Will climate change be more relentless than evolution?

Warm Fire

Fire in Kaibab National Forest, Arizona, USA. More frequent wildfires are just one way in which changing climate may exert selection on natural communities.

Ask any biologist what she considers the most urgently important example of adaptive evolution, and—even if she isn’t currently writing a grant proposal—she’ll probably mention global climate change. More than a century of pumping greenhouse gasses into Earth’s atmosphere has effectively set up a planet-wide experiment in evolution to answer the question, what species will be able to adapt to a warmer world?

Based on the results of a new study by Ignacio Quintero and John J. Wiens, the answer may be, not many. (I first saw coverage of the paper in a piece over at Mother Jones, and when I looked up the original paper in Ecology Letters, I thought it looked worthy of discussion here at The Molecular Ecologist, for reasons that should become obvious.)

The technical evolutionary question here is to do with the rate of adaptive evolutionary change—that is, change in response to natural selection exerted by rising global temperatures. Quintero and Wiens make some clever use of publicly available data to attempt to answer that question. Those data are phylogenetic relationships, locations, and climate:

  • The authors identified 17 vertebrate families for which there are published, time-calibrated, DNA sequence-based estimates of ancestral relationships among at least 80% of the known species in each family.
  • Then, they selected 270 pairs of sister species across all these different families, and paired the phylogenies with data on locations where the species in each pair were known to have been collected, from public databases like GBIF.
  • Using that location data, they queried the WorldClim database for average, maximum, and minimum annual temperature and average annual rainfall, rainfall in the wettest quarter, and rainfall in the driest quarter at all the locations where each species was found.

From the climate data for each species in the pairs, Quintero and Weins then conducted ancestral state reconstruction across the family-level phylogenies to estimate the climate conditions occupied by the common ancestor of each species pair. (Note that, although this applies some of the same tools and data resources, it’s not the same thing as formal niche model estimation—for many species, there wasn’t enough data do do that anyway.) Then, the authors estimated the rate of each species’ evolution in response to climate change as the difference between the conditions currently occupied by that species and the estimated conditions occupied by its ancestor, divided by the estimated time elapsed between modern species and reconstructed ancestor. So, to steal the example from the paper “… a species that differed by 5ºC from its reconstructed ancestor that is 5 million years (Myr) old would have a rate of 1ºC/Myr.”

To estimate the rate of evolutionary change necessary to respond to global warming, the authors used projections for climate conditions in the years 2080–2099, at locations currently occupied by each species, dividing the difference between that period and current conditions by 90 years. The comparison of that set of rates to the ones estimated from the phylogenies was stark:

Figure 1, Quintero and Wiens (2013)

Figure 1 from Quintero and Wiens (2013). For each vertebrate family, gray boxplots give the range of estimated rates of past change in climate (left to right, annual minimum temperature, annual mean temperature, and annual maximum temperature) and red boxplots give the estimated rate of change necessary to keep up with 90 years of global warming. Note that the y-axis is log-scaled.

To keep up with projected climate change, Quintero and Wiens estimated that the species in their dataset would have to undergo adaptive change at from 10,000 to 100,000 times faster than the rates estimated in their evolutionary past. Their Figure 1 (above) illustrates this result with glaring, log-scaled clarity.

On the one hand, this is an elegant application of the fruits of modern molecular methods—time-calibrated phylogenetic estimates—to answer a legitimately important question. But I’m not sure it tells us as much as we might want to know about potential evolutionary responses to climate change.

First off, all of the families considered are tetrapod vertebrates. A key characteristic of four-legged animals is that they can, you know, move. If animal populations tend to shift their ranges in response to climate, and have generally been able to do so faster than climate change exerted new selective pressures on them, they may not often have needed to evolve faster than what Quintero and Wiens estimate. And, more importantly, they won’t have to evolve in response to the full range of climate change estimated to occur over the next century—they can, to some extent, move.

(Yes, some of the groups considered, like the frogs and salamanders, may be a lot less mobile than, say, the birds. I’d still like to see a similar analysis for organisms that can’t even hop out of the way of warmer temperatures. Like, ahem, plants.)

More importantly, here’s the thing about global warming: it’s changing the planetary climate, by scientific consensus, at an unprecedented rate. That means that any reasonably accurate reconstruction of past rates of evolutionary response to climate change will necessarily be slower than the projected rate of change between now and the year 2100. Will it necessarily be 100,000 times slower? That we can’t tell from Quintero and Wiens’ data.

Finally, I can’t help but think back to Stephen Jay Gould’s “paradox of the visibly irrelevant”—the point that studies of year-to-year evolutionary change find rates of change that are invariably faster (even tens of thousands of times faster!) than estimates based on fossil (and, here, DNA-based) reconstructions. If I had to choose which kind of study is more appropriate for making predictions about adaptation in response to rapid climate change over the next several decades, I’d choose the short-term studies without much hesitation.

At the scale of such short-term change, we run into a whole bunch of different complicating questions, most prominently whether natural populations have enough genetic variation to fuel an adaptive response to climate change. But, again, we can’t answer that question with this data.

To their credit, Quintero and Wiens acknowledge all of these issues. They particularly note that year-to-year studies of species shifting their ranges or adapting to climate change have so far suggested that they’re already having trouble keeping up. Even accounting for the rapid rate of projected change relative to past climate change and the possibility of more rapid evolution on shorter time scales, they argue, the Earth’s living populations will have to deal with a mighty big change in the near future—change more rapid than anything they or their ancestors have weathered:

Nevertheless, combining our (very coarse) climatic projections with current data for these species suggests that future conditions will be outside the current climatic niches of many species … especially for temperature for many tropical species. These species may require extensive dispersal or niche evolution to survive projected changes.

In other words, while this study may not provide a very precise prediction of specific species’ capacity to evolve out of the way of global change, it does give us plenty of reason for concern. On that point, we’re entirely in agreement.

Reference

Quintero I., Wiens J. J., 2013 Rates of projected climate change dramatically exceed past rates of climatic niche evolution among vertebrate species. Ecology Letters. doi: 10.1111/ele.12144.

RedditDiggMendeleyPocketShare and Enjoy

About Jeremy Yoder

Jeremy Yoder is a postdoctoral associate in the Department of Plant Biology at the University of Minnesota. He also blogs at Denim and Tweed and Nothing in Biology Makes Sense!, and tweets under the handle @jbyoder.
This entry was posted in phylogenetics, quantitative genetics and tagged . Bookmark the permalink.
  • Conor Savage O’Brien

    Good explanation of an interesting study. I’m a couple years out of date on the literature, but to my knowledge, evolution in response to changing temperatures has not been demonstrated. What has been shown is that timing of seasonal behaviors has evolved. Here’s a good article to start with, if you’re interested in further reading. Unfortunately my knowledge (and the field at large) is animal centric:

    Bradshaw & Holzafpel 2008, “Genetic response to rapid climate change: it’s seasonal timing that matters.”
    http://onlinelibrary.wiley.com/doi/10.1111/j.1365-294X.2007.03509.x/full

    • http://www.denimandtweed.com/ Jeremy Yoder

      As a matter of fact, I’ve been working on a manuscript on the subject of local adaptation to climate variation, so I’m reasonably up-to-speed (I’d like to think) on this literature—including this idea that it’s all about seasonal timing, which manifests in relation to plants in a lot of studies looking at genes in the flowering time pathway.

      Of course, in the context of any climactic study, dealing with something like temperature is often linked to seasonal timing, and so the “adaptation” that Quintero and Wiens are estimating may be about timing of, e.g., reproductive activity or migration rather than tolerance of heat per se.

  • Pingback: Climate change could leave Darwin far behind |

  • Pingback: Scanning the genome for local adaptation | The Molecular Ecologist

  • Pingback: The Molecular Ecologist: Climate’s a-changin’. Will the living world evolve to cope? | Jeremy Yoder